Determination of Enantiomeric Excess by zy ”C NMR Spectroscopy? V. N. zyxwvut Gogte* zyxwvutsrq National Chemical Laboratory, Pune-411 008, India R. K. Nanda Hindustan Antibiotics Ltd, Pimpri, Pune-411 018, India A. A. Natu and V. S. Pandit National Chemical Laboratory, Pune-411 008, India M. K. Sastry Hindustan Antibiotics Ltd, Pimpri, Pune-411 018, India Oxazolidine derivatives of p-amino alcohols such as ephedrine have been resolved by zyx -C NMR spedroscopy using Eu(hfc), as a chiral shift reagent. The method is quantitative in the determination of enantiomeric excess, and zyxwvutsrq is advantageous where ‘H NMR is of limited use owing, for example, to signiticant line broadening. INTRODUCTION The /3 -amino alcohol moiety constitutes the backbone of a large number of sympathomimetic drugs.’ Their optical purity is of prime importance in their bioactiv- ity. During the syntheses of optically active p-amino alcohol derivatives it was necessary to develop a method to determine their optical purity, in particular for compounds whose rotations were unknown. Of the several methods available for the determina- tion of optical purity, that based on NMR spectros- copy has many advantage^.^ Although the ‘H NMR method incorporating the use of chiral shift reagents (CLSR) was successfully used in the case of 6- aminomercaptan derivatives, such as penicillamine diacetonide; it could not be used for the zyxwvut 6- aminoalcohol derivatives synthesized by us. The use of the same techniques using 6.0mol equivalent YO of CLSR for their 1,3-oxazolidine derivatives (e.g. Fig. 1) again did not produce good results. The use of spin decoupling in conjunction with the above techniques did not reduce the complexity of the spectra. In spite of the limitations of the use of 13C NMR spectroscopy in solving such a problem,’ an attempt was made to use this technique with suitable modifica- tions. The proposed methodology was tried on a stan- dard compound, ephedrine (2), since its chemistry and absolute configuration have been well described in the literature.‘ * Author to whom correspondence should be addressed. t NCL Communication No. 3330. Part VI of the series ‘Optical Induction’, published by NCL. For Part V, see Ref. 1. EXPERIMENTAL I3C zyxwv FT NMR (completely proton noise decoupled) spectra were recorded on a Jeol FX- 100 spectrometer operating at 25.05 MHz and equipped with a Jeol 980 computer. Typical conditions for the measurements were as follows: spectral width 4000Hz, data points 16K, pulse width 6 zyxw ks, pulse repetition 3 s, reference standard TMS, exponential factor 7, temperature am- bient, concentration 150mg in 3ml of CDCl,, deuterium lock and tube size 10 mm. The accuracy of the spectral line positions was zyx 0.5Hz. The spectra were calibrated using the central line of the CDCI, signal as 77 ppm. Tris[3-(heptafluoropropylhydroxy- methy1ene)-d-camphorato]europium(III) [Eu(hfc),] (1) was used as the CLSR. The quantity of the reagent added was sufficient for the separation of any one pair of 13C signals to be resolved down to the baseline. Pure d- and l-ephedrine (Aldrich) (2A and 2B, re- spectively) were converted into their respective ox- azolidine derivatives (3,4-dimethyl-5-phenyl-1,3- oxazolidines) by a known procedure.’ RESULTS AND DISCUSSION The limitations in the use of 13C NMR for the deter- mination of enantiomeric excess (e.e.) were partly overcome because the compounds under consideration were diastereoisomeric, and it was expected that the difference in the relaxation time, and in the NOE, experienced by the diastereoisomeric carbons would CCC-0030-4921/84/0022-0624$02.50 624 ORGANIC MAGNETIC RESONANCE, VOL. 22, NO. 10, 1984 zyxwvuts 0 Wiley Heyden Ltd, 1984